TCP Encapsulation of IKE and IPsec PacketsELVIS-PLUSPO Box 81Moscow (Zelenograd)124460Russian Federation+7 495 276 0211svan@elvis.ruApple Inc.1 Infinite LoopCupertino, California 95014United States of Americatpauly@apple.com This document describes a method to transport Internet Key Exchange
Protocol (IKE) and IPsec packets over a TCP connection for traversing
network middleboxes that may block IKE negotiation over UDP. This
method, referred to as "TCP encapsulation", involves sending both IKE
packets for Security Association establishment and Encapsulating
Security Payload (ESP) packets over a TCP connection. This method is
intended to be used as a fallback option when IKE cannot be
negotiated over UDP.
TCP encapsulation for IKE and IPsec was defined in RFC 8229.
This document updates the specification for TCP encapsulation by including
additional clarifications obtained during implementation and deployment
of this method. This documents obsoletes RFC 8229.
The Internet Key Exchange Protocol version 2 (IKEv2) is a
protocol for establishing IPsec Security Associations (SAs), using
IKE messages over UDP for control traffic, and using Encapsulating
Security Payload (ESP) messages for encrypted data traffic.
Many network middleboxes that filter traffic on public hotspots block
all UDP traffic, including IKE and IPsec, but allow TCP connections
through because they appear to be web traffic. Devices on these
networks that need to use IPsec (to access private enterprise
networks, to route Voice over IP calls to carrier networks, or
because of security policies) are unable to establish IPsec SAs.
This document defines a method for encapsulating IKE control messages
as well as IPsec data messages within a TCP connection.
Using TCP as a transport for IPsec packets adds a third option to the
list of traditional IPsec transports:
Direct. Currently, IKE negotiations begin over UDP port 500. If
no Network Address Translation (NAT) device is detected between
the Initiator and the Responder, then subsequent IKE packets are
sent over UDP port 500, and IPsec data packets are sent
using ESP.UDP Encapsulation . If a NAT is detected between the
Initiator and the Responder, then subsequent IKE packets are sent
over UDP port 4500 with four bytes of zero at the start of the
UDP payload, and ESP packets are sent out over UDP port 4500.
Some peers default to using UDP encapsulation even when no NAT is
detected on the path, as some middleboxes do not support IP
protocols other than TCP and UDP.TCP Encapsulation. If the other two methods are not available or
appropriate, IKE negotiation packets as well as ESP packets can
be sent over a single TCP connection to the peer. Direct use of ESP or UDP encapsulation should be preferred by
IKE implementations due to performance concerns when using
TCP encapsulation (). Most implementations should use
TCP encapsulation only on networks where negotiation over UDP has
been attempted without receiving responses from the peer or if a
network is known to not support UDP. Encapsulating IKE connections within TCP streams is a common approach
to solve the problem of UDP packets being blocked by network
middleboxes. The specific goals of this document are as follows:To promote interoperability by defining a standard method of
framing IKE and ESP messages within TCP streams.To be compatible with the current IKEv2 standard without requiring
modifications or extensions.To use IKE over UDP by default to avoid the overhead of other
alternatives that always rely on TCP or Transport Layer Security
(TLS) .Some previous alternatives include:
Interworking Wireless LAN (IWLAN) uses IKEv2 to create secure
connections to cellular carrier networks for making voice calls
and accessing other network services over Wi-Fi networks. 3GPP has
recommended that IKEv2 and ESP packets be sent within a TLS
connection to be able to establish connections on restrictive
networks.
Various non-standard extensions to the Internet Security
Association and Key Management Protocol (ISAKMP) have been
deployed that send IPsec traffic over TCP or TCP-like packets.
Many proprietary VPN solutions use a combination of TLS and IPsec
in order to provide reliability. These often run on TCP port 443.
IKEv2 over TCP as described in is used to avoid UDP
fragmentation.
TCP encapsulation for IKE and IPsec was defined in .
This document updates the specification for TCP encapsulation by including
additional clarifications obtained during implementation and deployment
of this method.
This document distinguishes between the IKE peer that initiates TCP
connections to be used for TCP encapsulation and the roles of
Initiator and Responder for particular IKE messages. During the
course of IKE exchanges, the role of IKE Initiator and Responder may
swap for a given SA (as with IKE SA rekeys), while the Initiator of
the TCP connection is still responsible for tearing down the TCP
connection and re-establishing it if necessary. For this reason,
this document will use the term "TCP Originator" to indicate the IKE
peer that initiates TCP connections. The peer that receives TCP
connections will be referred to as the "TCP Responder". If an IKE SA
is rekeyed one or more times, the TCP Originator MUST remain the peer
that originally initiated the first IKE SA.
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT",
"RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted
as described in BCP 14 when, and only when,
they appear in all capitals, as shown here.
One of the main reasons to use TCP encapsulation is that UDP traffic
may be entirely blocked on a network. Because of this, support for
TCP encapsulation is not specifically negotiated in the IKE exchange.
Instead, support for TCP encapsulation must be pre-configured on both
the TCP Originator and the TCP Responder.Implementations MUST support TCP encapsulation on TCP port 4500,
which is reserved for IPsec NAT traversal.Beyond a flag indicating support for TCP encapsulation, the
configuration for each peer can include the following optional
parameters:Alternate TCP ports on which the specific TCP Responder listens
for incoming connections. Note that the TCP Originator may
initiate TCP connections to the TCP Responder from any local port.An extra framing protocol to use on top of TCP to further
encapsulate the stream of IKE and IPsec packets. See
for a detailed discussion.Since TCP encapsulation of IKE and IPsec packets adds overhead and
has potential performance trade-offs compared to direct or
UDP-encapsulated SAs (as described in ), implementations
SHOULD prefer ESP direct or UDP-encapsulated SAs over
TCP-encapsulated SAs when possible.
Like UDP encapsulation, TCP encapsulation uses the first four bytes
of a message to differentiate IKE and ESP messages. TCP
encapsulation also adds a 16-bit Length field that precedes every message
to define the boundaries of messages within a stream.
The value in this field is equal to the length of the original message
plus the length of the field itself, in octets. If the first 32 bits
of the message are zeros (a non-ESP marker), then the contents comprise an
IKE message. Otherwise, the contents comprise an ESP message.
Authentication Header (AH) messages are not supported for TCP
encapsulation.
Although a TCP stream may be able to send very long messages,
implementations SHOULD limit message lengths to typical UDP datagram
ESP payload lengths. The maximum message length is used as the
effective MTU for connections that are being encrypted using ESP, so
the maximum message length will influence characteristics of inner
connections, such as the TCP Maximum Segment Size (MSS).
Note that this method of encapsulation will also work for placing IKE
and ESP messages within any protocol that presents a stream
abstraction, beyond TCP.
The IKE header is preceded by a 16-bit Length field in network byte
order that specifies the length of the IKE message (including the
non-ESP marker) within the TCP stream. As with IKE over UDP
port 4500, a zeroed 32-bit non-ESP marker is inserted before the
start of the IKE header in order to differentiate the traffic from
ESP traffic between the same addresses and ports.
Length (2 octets, unsigned integer) - Length of the IKE packet,
including the Length field and non-ESP marker. The value in the Length
field MUST NOT be 0 or 1. The receiver MUST treat these values as
fatal errors and MUST close TCP connection.
The ESP header is preceded by a 16-bit Length field in network byte
order that specifies the length of the ESP packet within the TCP
stream.The Security Parameter Index (SPI) field in the ESP header
MUST NOT be a zero value.Length (2 octets, unsigned integer) - Length of the ESP packet,
including the Length field. The value in the Length
field MUST NOT be 0 or 1. The receiver MUST treat these values as
fatal errors and MUST close TCP connection.Each stream of bytes used for IKE and IPsec encapsulation MUST begin
with a fixed sequence of six bytes as a magic value, containing the
characters "IKETCP" as ASCII values. This value is intended to
identify and validate that the TCP connection is being used for TCP
encapsulation as defined in this document, to avoid conflicts with
the prevalence of previous non-standard protocols that used TCP
port 4500. This value is only sent once, by the TCP Originator only,
at the beginning of any stream of IKE and ESP messages.
If other framing protocols are used within TCP to further encapsulate
or encrypt the stream of IKE and ESP messages, the stream prefix must
be at the start of the TCP Originator's IKE and ESP message stream
within the added protocol layer (). Although some framing
protocols do support negotiating inner protocols, the stream prefix
should always be used in order for implementations to be as generic
as possible and not rely on other framing protocols on top of TCP.
TCP encapsulation is applicable only when it has been configured to
be used with specific IKE peers. If a Responder is configured to use
TCP encapsulation, it MUST listen on the configured port(s) in case
any peers will initiate new IKE sessions. Initiators MAY use TCP
encapsulation for any IKE session to a peer that is configured to
support TCP encapsulation, although it is recommended that Initiators
should only use TCP encapsulation when traffic over UDP is blocked.
Since the support of TCP encapsulation is a configured property, not
a negotiated one, it is recommended that if there are multiple IKE
endpoints representing a single peer (such as multiple machines with
different IP addresses when connecting by Fully Qualified Domain
Name, or endpoints used with IKE redirection), all of the endpoints
equally support TCP encapsulation.
If TCP encapsulation is being used for a specific IKE SA, all
messages for that IKE SA and its Child SAs MUST be sent over a TCP
connection until the SA is deleted or IKEv2 Mobility and Multihoming
(MOBIKE) is used to change the SA endpoints and/or the encapsulation
protocol. See for more details on using MOBIKE to
transition between encapsulation modes.
Since UDP is the preferred method of transport for IKE messages,
implementations that use TCP encapsulation should have an algorithm
for deciding when to use TCP after determining that UDP is unusable.
If an Initiator implementation has no prior knowledge about the
network it is on and the status of UDP on that network, it SHOULD
always attempt to negotiate IKE over UDP first. IKEv2 defines how to
use retransmission timers with IKE messages and, specifically,
IKE_SA_INIT messages . Generally, this means that the
implementation will define a frequency of retransmission and the
maximum number of retransmissions allowed before marking the IKE SA
as failed. An implementation can attempt negotiation over TCP once
it has hit the maximum retransmissions over UDP, or slightly before
to reduce connection setup delays. It is recommended that the
initial message over UDP be retransmitted at least once before
falling back to TCP, unless the Initiator knows beforehand that the
network is likely to block UDP.
When switching from UDP to TCP, a new IKE_SA_INIT exchange MUST be
initiated with new Initiator's SPI and with recalculated content of
NAT_DETECTION_SOURCE_IP notification.
When the IKE Initiator uses TCP encapsulation, it will initiate a TCP
connection to the Responder using the configured TCP port. The first
bytes sent on the stream MUST be the stream prefix value ().
After this prefix, encapsulated IKE messages will negotiate the IKE
SA and initial Child SA . After this point, both
encapsulated IKE () and ESP ()
messages will be sent over the TCP connection. The TCP Responder MUST wait for the entire
stream prefix to be received on the stream before trying to parse out
any IKE or ESP messages. The stream prefix is sent only once, and
only by the TCP Originator.
In order to close an IKE session, either the Initiator or Responder
SHOULD gracefully tear down IKE SAs with DELETE payloads. Once the
SA has been deleted, the TCP Originator SHOULD close the TCP
connection if it does not intend to use the connection for another
IKE session to the TCP Responder. If the TCP connection is no more
associated with any active IKE SA, the TCP Responder MAY close the connection
to clean up resources if TCP Originator didn't close it within some reasonable period of time.
An unexpected FIN or a TCP Reset on the TCP connection may indicate a
loss of connectivity, an attack, or some other error. If a DELETE
payload has not been sent, both sides SHOULD maintain the state for
their SAs for the standard lifetime or timeout period. The TCP
Originator is responsible for re-establishing the TCP connection if
it is torn down for any unexpected reason. Since new TCP connections
may use different ports due to NAT mappings or local port allocations
changing, the TCP Responder MUST allow packets for existing SAs to be
received from new source ports.
A peer MUST discard a partially received message due to a broken
connection.
Whenever the TCP Originator opens a new TCP connection to be used for
an existing IKE SA, it MUST send the stream prefix first, before any
IKE or ESP messages. This follows the same behavior as the initial
TCP connection.
If a TCP connection is being used to resume a previous IKE session,
the TCP Responder can recognize the session using either the IKE SPI
from an encapsulated IKE message or the ESP SPI from an encapsulated
ESP message. If the session had been fully established previously,
it is suggested that the TCP Originator send an UPDATE_SA_ADDRESSES
message if MOBIKE is supported, or an informational message (a
keep-alive) otherwise.
The TCP Responder MUST NOT accept any messages for the existing IKE
session on a new incoming connection, unless that connection begins
with the stream prefix. If either the TCP Originator or TCP
Responder detects corruption on a connection that was started with a
valid stream prefix, it SHOULD close the TCP connection. The
connection can be determined to be corrupted if there are too many
subsequent messages that cannot be parsed as valid IKE messages or
ESP messages with known SPIs, or if the authentication check for an
ESP message with a known SPI fails. Implementations SHOULD NOT
tear down a connection if only a single ESP message has an unknown
SPI, since the SPI databases may be momentarily out of sync. If
there is instead a syntax issue within an IKE message, an
implementation MUST send the INVALID_SYNTAX notify payload and
tear down the IKE SA as usual, rather than tearing down the TCP
connection directly.
A TCP Originator SHOULD only open one TCP connection per IKE SA, over
which it sends all of the corresponding IKE and ESP messages. This
helps ensure that any firewall or NAT mappings allocated for the TCP
connection apply to all of the traffic associated with the IKE SA
equally.
Similarly, a TCP Responder SHOULD at any given time send packets for
an IKE SA and its Child SAs over only one TCP connection. It SHOULD
choose the TCP connection on which it last received a valid and
decryptable IKE or ESP message. In order to be considered valid for
choosing a TCP connection, an IKE message must be successfully
decrypted and authenticated, not be a retransmission of a previously
received message, and be within the expected window for IKE
message IDs. Similarly, an ESP message must pass authentication
checks and be decrypted, and must not be a replay of a previous
message.
Since a connection may be broken and a new connection re-established
by the TCP Originator without the TCP Responder being aware, a TCP
Responder SHOULD accept receiving IKE and ESP messages on both old
and new connections until the old connection is closed by the TCP
Originator. A TCP Responder MAY close a TCP connection that it
perceives as idle and extraneous (one previously used for IKE and ESP
messages that has been replaced by a new connection).
Multiple IKE SAs MUST NOT share a single TCP connection, unless one
is a rekey of an existing IKE SA, in which case there will
temporarily be two IKE SAs on the same TCP connection.
Section 2.1 of describes how IKEv2 deals with the unreliability
of the UDP protocol. In brief, the exchange Initiator is responsible for
retransmissions and must retransmit requests message until response
message is received. If no reply is received after several
retransmissions, the SA is deleted. The Responder never initiates retransmission,
but must send a response message again in case it receives a retransmitted request.
When IKEv2 uses a reliable transport protocol, like TCP, the retransmission rules are as follows:
The exchange Initiator SHOULD NOT retransmit request message; if
no response is received within some reasonable period of time, the
IKE SA is deleted.
If a new TCP connection for the IKE SA is established while the exchange
Initiator is waiting for a response, the Initiator MUST
retransmit its request over this connection and continue to wait for a response.
The exchange Responder does not change its behavior, but acts as
described in Section 2.1 of .
IKEv2 provides a DoS attack protection mechanism through Cookies, which
is described in Section 2.6 of . extends this
mechanism for protection against DDoS attacks by means of Client
Puzzles. Both mechanisms allow the Responder to avoid keeping state until
the Initiator proves its IP address is legitimate (and after solving a puzzle if required).
The connection-oriented nature of TCP transport brings additional
considerations for using these mechanisms.
In general, Cookies provide less value in case of TCP encapsulation,
since by the time a Responder receives the IKE_SA_INIT request, the TCP
session has already been established and the Initiator's IP address
has been verified. Moreover, a TCP/IP stack creates state once
a TCP SYN packet is received (unless SYN Cookies described in
are employed), which contradicts the statelessness of IKEv2 Cookies.
In particular, with TCP, an attacker is able to mount a SYN flooding DoS attack
which an IKEv2 Responder cannot prevent using stateless IKEv2 Cookies.
Thus, when using TCP encapsulation, it makes little sense to send Cookie requests without
Puzzles unless the Responder is concerned with a possibility of TCP
Sequence Number attacks (see for details). Puzzles, on
the other hand, still remain useful (and their use requires using Cookies).
The following considerations are applicable for using Cookie and
Puzzle mechanisms in case of TCP encapsulation:
the exchange Responder SHOULD NOT request a Cookie, with the
exception of Puzzles or in rare cases like preventing TCP Sequence
Number attacks.
if the Responder chooses to send Cookie request (possibly along
with Puzzle request), then the TCP connection that the IKE_SA_INIT
request message was received over SHOULD be closed after the Responder sends its reply
and no repeated requests are received within some short period of time
to keep the Responder stateless (see ). Note that the Responder MUST NOT
include the Initiator's TCP port into the Cookie
calculation (*), since the Cookie can be returned over a new
TCP connection with a different port.
the exchange Initiator acts as described in Section 2.6 of
and Section 7 of ,
i.e. using TCP encapsulation doesn't change the Initiator's behavior.
(*) Examples of Cookie calculation methods are given in Section 2.6
of and in Section 7.1.1.3 of
and they don't include transport protocol ports. However these examples are given
for illustrative purposes, since Cookie generation algorithm is a
local matter and some implementations might include port numbers,
that won't work with TCP encapsulation. Note also that these
examples include the Initiator's IP address in Cookie calculation.
In general this address may change between two initial requests (with and without Cookies).
This may happen due to NATs, since NATs have more freedom to change change source IP addresses for new
TCP connections than for UDP. In such cases cookie verification might fail.
There is a trade-off in choosing the period of time after which
TCP connection is closed. If it is too short, then the proper Initiator
which repeats its request would need to re-establish the TCP connection
introducing additional delay. On the other hand, if it is too long, then
the Responder's resources would be wasted in case the Initiator never comes back.
This document doesn't specify the duration of time, because it doesn't affect interoperability,
but it is believed that 5-10 seconds is a good compromise. Note also, that if the Responder requests
the Initiator to solve a puzzle, then the Responder can estimate how long it would take the Initiator
to find a solution and adjust the time interval accordingly.
Section 2.21.1 of describes how error notifications
are handled in the IKE_SA_INIT exchange. In particular, it is advised
that the Initiator should not act immediately after receiving error
notification and should instead wait some time for valid response,
since the IKE_SA_INIT messages are completely unauthenticated. This
advice does not apply equally in case of TCP encapsulation. If the
Initiator receives a response message over TCP, then either this
message is genuine and was sent by the peer, or the TCP session was
hijacked and the message is forged. In this latter case, no genuine
messages from the Responder will be received.
Thus, in case of TCP encapsulation, an Initiator SHOULD NOT wait for
additional messages in case it receives error notification from the
Responder in the IKE_SA_INIT exchange.
If in the IKE_SA_INIT exchange the Responder returns an error notification that implies
a recovery action from the Initiator (such as INVALID_KE_PAYLOAD
or INVALID_MAJOR_VERSION, see Section 2.21.1 of )
then the Responder SHOULD NOT close the TCP connection immediately, in anticipation
that the Initiator will repeat the request with corrected parameters.
See also .
When negotiating over UDP port 500, IKE_SA_INIT packets include
NAT_DETECTION_SOURCE_IP and NAT_DETECTION_DESTINATION_IP payloads to
determine if UDP encapsulation of IPsec packets should be used.
These payloads contain SHA-1 digests of the SPIs, IP addresses, and
ports as defined in . IKE_SA_INIT packets sent on a TCP
connection SHOULD include these payloads with the same content as
when sending over UDP and SHOULD use the applicable TCP ports when
creating and checking the SHA-1 digests.
If a NAT is detected due to the SHA-1 digests not matching the
expected values, no change should be made for encapsulation of
subsequent IKE or ESP packets, since TCP encapsulation inherently
supports NAT traversal. However, for the transport mode IPsec SAs, implementations
need to handle TCP and UDP packet checksum fixup during decapsulation,
as defined for UDP encapsulation in .
Implementations MAY use the information that
a NAT is present to influence keep-alive timer values.
Encapsulating IKE and IPsec inside of a TCP connection can impact the
strategy that implementations use to detect peer liveness and to
maintain middlebox port mappings. Peer liveness should be checked
using IKE informational packets .
In general, TCP port mappings are maintained by NATs longer than UDP
port mappings, so IPsec ESP NAT keep-alives SHOULD NOT be
sent when using TCP encapsulation. Any implementation using TCP
encapsulation MUST silently drop incoming NAT keep-alive packets
and not treat them as errors. NAT keep-alive packets over a
TCP-encapsulated IPsec connection will be sent as an ESP message with
a one-octet-long payload with the value 0xFF.
Note that, depending on the configuration of TCP and TLS on the
connection, TCP keep-alives and TLS keep-alives
may be used. These MUST NOT be used as indications of IKE peer
liveness, for which purpose the standard IKEv2 mechanism of exchanging empty INFORMATIONAL messages is used
(see Section 1.4 of ).
Using TCP encapsulation affects some aspects of IPsec SA processing.
Section 8.1 of requires all tunnel mode IPsec SAs to
be able to copy the Don't Fragment (DF) bit from inner IP header to
the outer (tunnel) one. With TCP encapsulation this is generally
not possible, because TCP/IP stack manages DF bit in the outer IP
header, and usually the stack ensures that the DF bit is set for TCP
packets to avoid IP fragmentation.
The other feature that is less applicable with TCP encapsulation is an
ability to split traffic of different QoS classes into different
IPsec SAs, created by a single IKE SA. In this case the
Differentiated Services Code Point (DSCP) field is usually copied
from the inner IP header to the outer (tunnel) one, ensuring that
IPsec traffic of each SA receives the corresponding level of service.
With TCP encapsulation all IPsec SAs created by a single IKE SA will
share a single TCP connection and thus will receive the same level of
service (see ). If this functionality is needed,
implementations should create several IKE SAs each over separate TCP connection
and assign a corresponding DSCP value to each of them.
Besides, TCP encapsulation of IPsec packets may have implications
on performance of the encapsulated traffic. Performance considerations
are discussed in .
The MOBIKE protocol, which allows SAs to migrate between IP
addresses, is defined in , and further clarifies
the details of the protocol. When an IKE session that has negotiated MOBIKE is
transitioning between networks, the Initiator of the transition may
switch between using TCP encapsulation, UDP encapsulation, or no
encapsulation. Implementations that implement both MOBIKE and TCP
encapsulation within the same connection configuration
MUST support dynamically enabling and disabling TCP
encapsulation as interfaces change.
When a MOBIKE-enabled Initiator changes networks, the
INFORMATIONAL exchange with the UPDATE_SA_ADDRESSES notification SHOULD be initiated
first over UDP before attempting over TCP. If there is a response to the
request sent over UDP, then the ESP packets should be sent directly over IP or over UDP port 4500
(depending on if a NAT was detected), regardless of if a connection on a previous
network was using TCP encapsulation. If no response is received within a certain period of time after
several retransmissions, the Initiator ought to change its transport for this exchange from
UDP to TCP and resend the request message. A new INFORMATIONAL exchange MUST
NOT be started in this situation. If the Responder only responds to the request sent over TCP, then
the ESP packets should be sent over the TCP connection, regardless of
if a connection on a previous network did not use TCP encapsulation.
If the TCP transport was used for the previous network connection, the old TCP
connection SHOULD be closed by the Initiator once MOBIKE finishes migration
to a new connection (either TCP or UDP).
Since switching from UDP to TCP can happen during a single
INFORMATIONAL message exchange, the content of the NAT_DETECTION_SOURCE_IP
notification will in most cases be incorrect (since UDP and TCP source ports
will most likely be different), and the peer may incorrectly detect
the presence of a NAT. Section 3.5 of states that
a new INFORMATIONAL exchange with the UPDATE_SA_ADDRESSES notify is initiated
in case the address (or transport) is changed while waiting for a response.
Section 3.5 of also states that
once an IKE SA is switched to a new IP address, all outstanding requests in this SA
are immediately retransmitted using this address. See also .
The MOBIKE protocol defines the NO_NATS_ALLOWED notification that can be
used to detect the presence of NAT between peer and to refuse to
communicate in this situation. In case of TCP the NO_NATS_ALLOWED
notification SHOULD be ignored because TCP generally has no problems
with NAT boxes.
Section 3.7 of describes an additional optional step in the
process of changing IP addresses called Return Routability Check. It
is performed by Responders in order to be sure that the new
initiator's address is in fact routable. In case of TCP
encapsulation this check has little value, since TCP handshake proves
routability of the TCP Originator's address. So, in case of TCP
encapsulation the Return Routability Check SHOULD NOT be performed.
A redirect mechanism for IKEv2 is defined in . This mechanism
allows security gateways to redirect clients to another gateway
either during IKE SA establishment or after session setup. If a
client is connecting to a security gateway using TCP and
then is redirected to another security gateway, the client
needs to reset its transport selection. In other words, the client
MUST again try first UDP and then fall back to TCP while establishing
a new IKE SA, regardless of the transport of the SA the redirect
notification was received over (unless the client's configuration
instructs it to instantly use TCP for the gateway it is redirected
to).
Session resumption for IKEv2 is defined in . Once an IKE SA is
established, the server creates a resumption ticket where information
about this SA is stored, and transfers this ticket to the client.
The ticket may be later used to resume the IKE SA after it is deleted.
In the event of resumption the client presents the ticket in a new
exchange, called IKE_SESSION_RESUME. Some parameters in the new SA
are retrieved from the ticket and others are re-negotiated (more details
are given in Section 5 of ).
Since network conditions may change while the client is incative,
the fact that TCP encapsulation was used in an old SA SHOULD NOT affect which transport
is used during session resumption. In other words, the transport should be
selected as if the IKE SA is being created from scratch.
defines a support for High Availability in IKEv2.
In case of cluster failover, a new active node
must immediately initiate a special INFORMATION exchange containing the
IKEV2_MESSAGE_ID_SYNC notification, which instructs the client to
skip some number of Message IDs that might not be synchronized yet
between nodes at the time of failover.
Synchronizing states when using TCP encapsulation is much harder than
when using UDP; doing so requires access to TCP/IP stack internals, which is
not always available from an IKE/IPsec implementation. If a cluster
implementation doesn't synchronize TCP states between nodes, then
after failover event the new active node will not have any TCP
connection with the client, so the node cannot initiate the
INFORMATIONAL exchange as required by . Since the cluster
usually acts as TCP Responder, the new active node cannot re-
establish TCP connection, since only the TCP Originator can do it.
For the client, the cluster failover event may remain
undetected for long time if it has no IKE or ESP traffic to send. Once
the client sends an ESP or IKEv2 packet, the cluster node will reply
with TCP RST and the client (as TCP Originator) will reestablish the TCP
connection so that the node will be able to initiate the
INFORMATIONAL exchange informing the client about the cluster
failover.
This document makes the following recommendation: if support for High
Availability in IKEv2 is negotiated and TCP transport is used,
a client that is a TCP Originator SHOULD periodically send
IKEv2 messages (e.g. by initiating liveness check exchange) whenever
there is no IKEv2 or ESP traffic. This differs from the
recommendations given in Section 2.4 of in the following:
the liveness check should be periodically performed even if the
client has nothing to send over ESP. The frequency of sending such
messages should be high enough to allow quick detection and restoring
of broken TCP connection.
IKE message fragmentation is not required when using TCP
encapsulation, since a TCP stream already handles the fragmentation
of its contents across packets. Since fragmentation is redundant in
this case, implementations might choose to not negotiate IKE
fragmentation. Even if fragmentation is negotiated, an
implementation SHOULD NOT send fragments when going over a TCP
connection, although it MUST support receiving fragments.
If an implementation supports both MOBIKE and IKE fragmentation, it
SHOULD negotiate IKE fragmentation over a TCP-encapsulated session in
case the session switches to UDP encapsulation on another network.
Many security networking devices, such as firewalls or intrusion
prevention systems, network optimization/acceleration devices, and
NAT devices, keep the state of sessions that traverse through them.
These devices commonly track the transport-layer and/or application-
layer data to drop traffic that is anomalous or malicious in nature.
While many of these devices will be more likely to pass
TCP-encapsulated traffic as opposed to UDP-encapsulated traffic, some
may still block or interfere with TCP-encapsulated IKE and IPsec
traffic.
A network device that monitors the transport layer will track the
state of TCP sessions, such as TCP sequence numbers. TCP
encapsulation of IKE should therefore use standard TCP behaviors to
avoid being dropped by middleboxes.
Several aspects of TCP encapsulation for IKE and IPsec packets may
negatively impact the performance of connections within a tunnel-mode
IPsec SA. Implementations should be aware of these performance
impacts and take these into consideration when determining when to
use TCP encapsulation. Implementations SHOULD favor using direct ESP
or UDP encapsulation over TCP encapsulation whenever possible.
If the outer connection between IKE peers is over TCP, inner TCP
connections may suffer negative effects from using TCP within TCP.
Running TCP within TCP is discouraged, since the TCP algorithms
generally assume that they are running over an unreliable datagram
layer.
If the outer (tunnel) TCP connection experiences packet loss, this
loss will be hidden from any inner TCP connections, since the outer
connection will retransmit to account for the losses. Since the
outer TCP connection will deliver the inner messages in order, any
messages after a lost packet may have to wait until the loss is
recovered. This means that loss on the outer connection will be
interpreted only as delay by inner connections. The burstiness of
inner traffic can increase, since a large number of inner packets may
be delivered across the tunnel at once. The inner TCP connection may
interpret a long period of delay as a transmission problem,
triggering a retransmission timeout, which will cause spurious
retransmissions. The sending rate of the inner connection may be
unnecessarily reduced if the retransmissions are not detected as
spurious in time.
The inner TCP connection's round-trip-time estimation will be
affected by the burstiness of the outer TCP connection if there are
long delays when packets are retransmitted by the outer TCP
connection. This will make the congestion control loop of the inner
TCP traffic less reactive, potentially permanently leading to a lower
sending rate than the outer TCP would allow for.
TCP-in-TCP can also lead to increased buffering, or bufferbloat.
This can occur when the window size of the outer TCP connection is
reduced and becomes smaller than the window sizes of the inner TCP
connections. This can lead to packets backing up in the outer TCP
connection's send buffers. In order to limit this effect, the outer
TCP connection should have limits on its send buffer size and on the
rate at which it reduces its window size.
Note that any negative effects will be shared between all flows going
through the outer TCP connection. This is of particular concern for
any latency-sensitive or real-time applications using the tunnel. If
such traffic is using a TCP-encapsulated IPsec connection, it is
recommended that the number of inner connections sharing the tunnel
be limited as much as possible.
Since ESP is an unreliable protocol, transmitting ESP packets over a
TCP connection will change the fundamental behavior of the packets.
Some application-level protocols that prefer packet loss to delay
(such as Voice over IP or other real-time protocols) may be
negatively impacted if their packets are retransmitted by the TCP
connection due to packet loss.
Quality-of-Service (QoS) markings, such as the Differentiated
Services Code Point (DSCP) and Traffic Class, should be used with
care on TCP connections used for encapsulation. Individual packets
SHOULD NOT use different markings than the rest of the connection,
since packets with different priorities may be routed differently and
cause unnecessary delays in the connection.
A TCP connection used for IKE encapsulation SHOULD negotiate its MSS
in order to avoid unnecessary fragmentation of packets.
Since there is not a one-to-one relationship between outer IP packets
and inner ESP/IP messages when using TCP encapsulation, the markings
for Explicit Congestion Notification (ECN) cannot be simply
mapped. However, any ECN Congestion Experienced (CE) marking on
inner headers should be preserved through the tunnel.
Implementations SHOULD follow the ECN compatibility mode for tunnel
ingress as described in . In compatibility mode, the outer
tunnel TCP connection marks its packet headers as not ECN-capable.
If upon egress, the arriving outer header is marked with CE, the
implementation will drop the inner packet, since there is not a
distinct inner packet header onto which to translate the ECN
markings.
IKE Responders that support TCP encapsulation may become vulnerable
to new Denial-of-Service (DoS) attacks that are specific to TCP, such
as SYN-flooding attacks. TCP Responders should be aware of this
additional attack surface.
TCP Responders should be careful to ensure that (1) the stream prefix
"IKETCP" uniquely identifies incoming streams as streams that use the
TCP encapsulation protocol and (2) they are not running any other
protocols on the same listening port (to avoid potential conflicts).
Attackers may be able to disrupt the TCP connection by sending
spurious TCP Reset packets. Therefore, implementations SHOULD make
sure that IKE session state persists even if the underlying TCP
connection is torn down.
If MOBIKE is being used, all of the security considerations outlined
for MOBIKE apply .
Similarly to MOBIKE, TCP encapsulation requires a TCP Responder to
handle changes to source address and port due to network or
connection disruption. The successful delivery of valid IKE or ESP
messages over a new TCP connection is used by the TCP Responder to
determine where to send subsequent responses. If an attacker is able
to send packets on a new TCP connection that pass the validation
checks of the TCP Responder, it can influence which path future
packets will take. For this reason, the validation of messages on
the TCP Responder must include decryption, authentication, and replay
checks.
TCP port 4500 is already allocated to IPsec for NAT traversal. This
port SHOULD be used for TCP-encapsulated IKE and ESP as described in
this document.
This document updates the reference for TCP port 4500 from RFC 8229 to itself:
This section provides recommendations on how to use TLS in addition
to TCP encapsulation.
When using TCP encapsulation, implementations may choose to use TLS 1.2
or TLS 1.3 on the TCP connection
to be able to traverse middleboxes, which may otherwise block the traffic.
If a web proxy is applied to the ports used for the TCP connection
and TLS is being used, the TCP Originator can send an HTTP CONNECT
message to establish an SA through the proxy .
The use of TLS should be configurable on the peers, and may be used
as the default when using TCP encapsulation or may be used as a
fallback when basic TCP encapsulation fails. The TCP Responder may
expect to read encapsulated IKE and ESP packets directly from the TCP
connection, or it may expect to read them from a stream of TLS data
packets. The TCP Originator should be pre-configured to use TLS
or not when communicating with a given port on the TCP Responder.
When new TCP connections are re-established due to a broken
connection, TLS must be renegotiated. TLS session resumption is
recommended to improve efficiency in this case.
The security of the IKE session is entirely derived from the IKE
negotiation and key establishment and not from the TLS session (which
in this context is only used for encapsulation purposes); therefore,
when TLS is used on the TCP connection, both the TCP Originator and
the TCP Responder SHOULD allow the NULL cipher to be selected for
performance reasons. Note, that TLS 1.3 only supports AEAD algorithms
and at the time of writing this document there was no recommended cipher suite
for TLS 1.3 with the NULL cipher.
Implementations should be aware that the use of TLS introduces
another layer of overhead requiring more bytes to transmit a given
IKE and IPsec packet. For this reason, direct ESP, UDP
encapsulation, or TCP encapsulation without TLS should be preferred
in situations in which TLS is not required in order to traverse
middleboxes.
IP_R:Port_R)
TcpSyn ---------->
<---------- TcpSyn,Ack
TcpAck ---------->
2) --------------------- TLS Session ---------------------
ClientHello ---------->
ServerHello
{EncryptedExtensions}
{Certificate*}
{CertificateVerify*}
<---------- {Finished}
{Finished} ---------->
3) ---------------------- Stream Prefix --------------------
"IKETCP" ---------->
4) ----------------------- IKE Session ---------------------
Length + Non-ESP Marker ---------->
IKE_SA_INIT
HDR, SAi1, KEi, Ni,
[N(NAT_DETECTION_*_IP)]
<------ Length + Non-ESP Marker
IKE_SA_INIT
HDR, SAr1, KEr, Nr,
[N(NAT_DETECTION_*_IP)]
Length + Non-ESP Marker ---------->
first IKE_AUTH
HDR, SK {IDi, [CERTREQ]
CP(CFG_REQUEST), IDr,
SAi2, TSi, TSr, ...}
<------ Length + Non-ESP Marker
first IKE_AUTH
HDR, SK {IDr, [CERT], AUTH,
EAP, SAr2, TSi, TSr}
Length + Non-ESP Marker ---------->
IKE_AUTH + EAP
repeat 1..N times
<------ Length + Non-ESP Marker
IKE_AUTH + EAP
Length + Non-ESP Marker ---------->
final IKE_AUTH
HDR, SK {AUTH}
<------ Length + Non-ESP Marker
final IKE_AUTH
HDR, SK {AUTH, CP(CFG_REPLY),
SA, TSi, TSr, ...}
-------------- IKE and IPsec SAs Established ------------
Length + ESP Frame ---------->
]]>The client establishes a TCP connection with the server on
port 4500 or on an alternate pre-configured port that the server
is listening on.
If configured to use TLS, the client initiates a TLS handshake.
During the TLS handshake, the server SHOULD NOT request the
client's certificate, since authentication is handled as part of
IKE negotiation.The client sends the stream prefix for TCP-encapsulated IKE
() traffic to signal the beginning of IKE negotiation.The client and server establish an IKE connection. This example
shows EAP-based authentication, although any authentication type
may be used.
INFORMATIONAL
HDR, SK {[N,] [D,]
[CP,] ...}
<------ Length + Non-ESP Marker
INFORMATIONAL
HDR, SK {[N,] [D,]
[CP], ...}
2) --------------------- TLS Session ---------------------
close_notify ---------->
<---------- close_notify
3) -------------------- TCP Connection -------------------
TcpFin ---------->
<---------- Ack
<---------- TcpFin
Ack ---------->
-------------------- IKE SA Deleted -------------------
]]>The client and server exchange informational messages to notify
IKE SA deletion.The client and server negotiate TLS session deletion using TLS
CLOSE_NOTIFY.The TCP connection is torn down.The deletion of the IKE SA should lead to the disposal of the
underlying TLS and TCP state. IP_R:Port_R)
TcpSyn ---------->
<---------- TcpSyn,Ack
TcpAck ---------->
2) --------------------- TLS Session ---------------------
ClientHello ---------->
ServerHello
{EncryptedExtensions}
<---------- {Finished}
{Finished} ---------->
3) ---------------------- Stream Prefix --------------------
"IKETCP" ---------->
4) <---------------------> IKE/ESP Flow <------------------>
]]>If a previous TCP connection was broken (for example, due to a
TCP Reset), the client is responsible for re-initiating the TCP
connection. The TCP Originator's address and port (IP_I and
Port_I) may be different from the previous connection's address
and port.
The client SHOULD attempt TLS session resumption if it
has previously established a session with the server.
After TCP and TLS are complete, the client sends the stream
prefix for TCP-encapsulated IKE traffic ().
The IKE and ESP packet flow can resume. If MOBIKE is being used,
the Initiator SHOULD send an UPDATE_SA_ADDRESSES message.
IP_R:UDP500)
1) ----------------- IKE_SA_INIT Exchange -----------------
(IP_I1:UDP4500 -> IP_R:UDP4500)
Non-ESP Marker ----------->
Initial IKE_AUTH
HDR, SK { IDi, CERT, AUTH,
CP(CFG_REQUEST),
SAi2, TSi, TSr,
N(MOBIKE_SUPPORTED) }
<----------- Non-ESP Marker
Initial IKE_AUTH
HDR, SK { IDr, CERT, AUTH,
EAP, SAr2, TSi, TSr,
N(MOBIKE_SUPPORTED) }
<------------------ IKE SA Establishment --------------->
2) ------------ MOBIKE Attempt on New Network --------------
(IP_I2:UDP4500 -> IP_R:UDP4500)
Non-ESP Marker ----------->
INFORMATIONAL
HDR, SK { N(UPDATE_SA_ADDRESSES),
N(NAT_DETECTION_SOURCE_IP),
N(NAT_DETECTION_DESTINATION_IP) }
3) -------------------- TCP Connection -------------------
(IP_I2:Port_I -> IP_R:Port_R)
TcpSyn ----------->
<----------- TcpSyn,Ack
TcpAck ----------->
4) --------------------- TLS Session ---------------------
ClientHello ---------->
ServerHello
{EncryptedExtensions}
{Certificate*}
{CertificateVerify*}
<---------- {Finished}
{Finished} ---------->
5) ---------------------- Stream Prefix --------------------
"IKETCP" ---------->
6) ------------ Retransmit Message from step 2 -------------
Length + Non-ESP Marker ----------->
INFORMATIONAL
HDR, SK { N(UPDATE_SA_ADDRESSES),
N(NAT_DETECTION_SOURCE_IP),
N(NAT_DETECTION_DESTINATION_IP) }
<------- Length + Non-ESP Marker
HDR, SK { N(NAT_DETECTION_SOURCE_IP),
N(NAT_DETECTION_DESTINATION_IP) }
7) ---- New Exchange with recalculated NAT_DETECTION_* ----
Length + Non-ESP Marker ----------->
INFORMATIONAL
HDR, SK { N(UPDATE_SA_ADDRESSES),
N(NAT_DETECTION_SOURCE_IP),
N(NAT_DETECTION_DESTINATION_IP) }
<------- Length + Non-ESP Marker
HDR, SK { N(NAT_DETECTION_SOURCE_IP),
N(NAT_DETECTION_DESTINATION_IP) }
8) <---------------------> IKE/ESP Flow <------------------>
]]>During the IKE_SA_INIT exchange, the client and server exchange
MOBIKE_SUPPORTED notify payloads to indicate support for MOBIKE.
The client changes its point of attachment to the network and
receives a new IP address. The client attempts to re-establish
the IKE session using the UPDATE_SA_ADDRESSES notify payload, but
the server does not respond because the network blocks UDP
traffic.
The client brings up a TCP connection to the server in order to
use TCP encapsulation.
The client initiates a TLS handshake with the server.The client sends the stream prefix for TCP-encapsulated IKE
traffic ().
The client sends the UPDATE_SA_ADDRESSES notify payload on the
TCP-encapsulated connection. Note that this IKE message is the
same as the one sent over UDP in step 2; it should have the same
message ID and contents.
Once the client receives a response on the
TCP-encapsulated connection, it immediately starts a new INFORMATIONAL
exchange with an UPDATE_SA_ADDRESSES notify payload and a recalculated
NAT_DETECTION_SOURCE_IP notify payload in order to get correct information about the presence
of NATs.
The IKE and ESP packet flow can resume.Thanks to the original authors of RFC 8229, Tommy Pauly, Samy Touati, and Ravi Mantha.
Since this document updates and obsoletes RFC 8229, most of its text was borrowed
from the original document.
The following people provided valuable feedback and advices while preparing RFC8229:
Stuart Cheshire, Delziel Fernandes, Yoav Nir, Christoph Paasch, Yaron
Sheffer, David Schinazi, Graham Bartlett, Byju Pularikkal, March Wu,
Kingwel Xie, Valery Smyslov, Jun Hu, and Tero Kivinen. Special
thanks to Eric Kinnear for his implementation work.
The authors would like to thank Tero Kivinen and Paul Wouters for their
valuable comments while preparing this document.